Adeno-associated viruses (AAVs) are naturally defective human parvoviruses that are being developed as delivery systems (vectors) for gene therapies to treat diabetes and many other chronic diseases. AAVs require co-infection with a helper virus, usually an adenovirus or herpesvirus, for efficient productive infection. In non-dividing cells, in the absence of helper virus, AAV DNA can persist episomally (i.e. without integrating into host chromosomes). Persistent infection of dividing cells requires integration into host chromosomes. The genome of the best studied AAV, AAV2, integrates into the human genome with a strong preference (70%-90% of the time) for a 4 kb region of human chromosome 19 (the only example of site-specific integration in a mammalian virus system). Preferential integration is mediated by protein products of the AAV2 rep gene. DNA of AAV-based gene delivery vectors that lack the rep gene can also persist episomally in non-dividing cells and integrate, but the integration is much less specific.[unreadable] Pancreatic islets are important targets for diabetes gene therapies. Although several diabetes model systems involve rats, transduction of rat pancreatic islets by AAV vectors has not been well characterized. Since the AAV packaging signals are highly conserved, it is possible to package a recombinant AAV2 genome, containing a marker or therapeutic gene (transgene), in the capsids of multiple AAV types (pseudotyping). We performed a head-to-head comparison of the same AAV2-based genome, encoding green fluorescent protein, packaged into AAV2, AAV5 or AAV8 capsids, for their abilities to transduce rat or mouse pancreatic islets. The islets were also infected with helper adenovirus type 5, which has been shown to speed AAV vector transgene expression. The AAV5-encapsidated vector worked the best. Flow cytometry analysis of rat islet cells that were dissociated 6 days after transduction indicated AAV-mediated gene expression in approximately 10% of rat islet cells overall and almost 12% of the insulin-producing beta islet cells, when the AAV5 capsid was used. We found that the AAV8-encapsidated vector had a higher dependence on the helper virus multiplicity of infection than the AAV5-encapsidated vector. We were also able to demonstrate transduction of human islets by an AAV8-encapsidated vector.[unreadable] Pancreatic islet transplantation has shown some modest success in the treatment of type 1 diabetes. Progress has been hindered by several factors. First, transplanted islets are recognized as foreign and are rejected by the immune system. Second, the autoimmune response that killed the recipients original islets, causing the original type 1 diabetes, can destroy the donor islets. Third, there is a limited supply of donor islets. To get around the first two problems, patients are treated with drugs to repress the immune system. Many of these drugs have such severe side-effects that their long-term use is not tolerable for many patients. We and others have hypothesized that the introduction of genes encoding proteins that locally suppress the immune system may protect the islets while minimizing the side effects associated with systemic immune suppression. One candidate gene is transforming growth factor-beta 1 (TGF-beta 1). There has been vigorous debate over whether or not TGF-beta 1 is itself harmful to islet functions. In order to test this hypothesis, we produced a recombinant AAV vector carrying the TGF-beta 1 gene, with an AAV5 capsid. Rat islets transduced with this vector and infected with Ad5 were compared to islets that had been treated with Ad5 only. Insulin levels in the medium taken from transduced islets were not significantly different from levels in medium taken from islets that had only been infected with Ad5. However, there was more than a four-fold increase in TGF-beta 1 levels in the conditioned medium from AAV-transduced/Ad5-infected islets, compared to islets treated only with Ad5. These results suggest that the AAV-TGF-beta 1 vector is able to transduce islets without altering the response of beta-islet cells to the 100 mg/dl glucose in the CMRL Medium-1066 used, while successfully expressing the TGF-beta 1 gene. These data concur with a previous report by another laboratory that used a recombinant adenovirus vector to deliver the TGF beta-1 gene to islets and showed no impact on insulin production in culture.[unreadable] The A3 adenosine receptor is involved in signal transduction that can prevent cell death. It has been shown to be an important factor in the modulation of kidney and muscle damage that can result from various insults. Our collaborators, based on protein structural data, have designed agonists that should specifically stimulate the function of either the wild-type or a modified version of the A3 receptor. As a first step in testing this hypothesis, we have used an AAV-based vector to express the wild-type A3 adenosine receptor in a mouse muscle cell line that does not normally express this gene. The goal of this line of experimentation is to test hypotheses about A3 receptor functions in vivo, under conditions which are unlikely to stimulate the functions of other adenosine receptors.